| 06:45.58 | seebs | Idle question, anyone happen to know of a way to cap framerate in rift? I know this is not a common requirement, given how low it usually is to begin with. :P |
| 06:50.17 | ZorbaTHut | seebs, if you're okay with a ton of CPU use, I think the latest Trion Dev Tools has a /framecap command - it just burns CPU to bring it down to that, though |
| 06:50.27 | seebs | Whee. |
| 06:50.50 | seebs | Well, the issue is, looking at upgrading the spousal laptop, and if we do, it will probably be able to run rift but may get a bit hot. |
| 06:51.04 | ZorbaTHut | ahh, that's a bit trickier |
| 06:51.10 | seebs | An option to tell rift that we'd rather get 15fps and not catch fire rather than 25 and fire would be very nice. |
| 06:51.18 | ZorbaTHut | I think most GPUs support clock speed control. usually that's used for overclocking, but you could underclock as well |
| 06:51.33 | seebs | That's a thought. ... but it'll be a Mac, so I don't know whether we have access. |
| 06:51.36 | ZorbaTHut | it'd probably save more power than an artificial fps cap, and also be more consistent |
| 06:51.39 | ZorbaTHut | mmm, good point |
| 06:51.50 | seebs | Long story short: If it comes down to either Boot Camp or not playing, spouse will probably prefer not playing. |
| 06:52.09 | seebs | I have verified that the game at least comes up under Crossover, but all my Macs are old and haven't the video hardware to play credibly. |
| 06:52.52 | seebs | I would be so much happier if rift had been cross-platform in the first place. Making code portable after the fact is usually a non-starter. |
| 06:53.06 | ZorbaTHut | cross-platform is, unfortunately, really really hard |
| 06:53.09 | seebs | Yeah. |
| 06:53.25 | seebs | Boot Camp has dramatically lowered the economic value of Mac ports of hard-to-port things. |
| 06:53.33 | ZorbaTHut | I'm writing a game engine right now that supports it, but it's more important for indie developers than it is for AAA studios |
| 06:53.45 | ZorbaTHut | Linux support increases your sales pretty significantly. |
| 06:53.51 | ZorbaTHut | Admittedly, most of those sales are Windows licenses. |
| 06:54.02 | seebs | I still believe that the Mac support has been a big part of WoW's success, just from network effects. |
| 06:54.20 | seebs | I know a single person who played on a Mac, and 8-10 accounts that were active in WoW for years because of that single person... |
| 06:54.28 | ZorbaTHut | unfortunately, I've heard a few stories from inside that Mac support has also been a big part of WoW's slow feature development |
| 06:55.18 | seebs | Interesting. I am mostly not doing game engine type stuff, so for my purposes, platform is about 99% irrelevant, since nothing I write talks to platforms anyway. |
| 06:55.44 | ZorbaTHut | I'm rigging up my rapid development engine with a lot of its cruft removed. gonna be so nice once it's done |
| 06:55.52 | seebs | Cool. |
| 06:56.02 | ZorbaTHut | https://github.com/zorbathut/glorp/tree/glorp2 wheeee :V |
| 06:56.11 | ZorbaTHut | anyway, off to watch ponies with the gf :) |
| 06:56.14 | seebs | ... actually, that's not quite true, I have a large hunk of excessively unportable code. |
| 06:56.17 | seebs | Yay ponies! Have fun! |
| 06:56.55 | seebs | I have a thing that intercepts Unixy syscalls and pretends to have root privs. I ported it from Linux to OS X largely because I'm unable to accept "that is stupid and cannot work" as an answer. |
| 07:03.28 | seebs | oooh |
| 07:03.32 | seebs | I just had the awesomest idea. |
| 07:03.56 | seebs | Which is a multi-pane mail reader UI for the rift mailbox. Just for the Looks Cool. |
| 07:50.56 | *** join/#riftuidev Torhal (~callahan@c-68-62-216-165.hsd1.al.comcast.net) |
| 07:50.56 | *** mode/#riftuidev [+v Torhal] by ChanServ |
| 08:18.16 | *** join/#riftuidev lXl (~Main@cpe-173-172-35-32.tx.res.rr.com) |
| 08:47.17 | ZorbaTHut | I'm quite curious - what did you need to fake root privs for? |
| 08:47.44 | seebs | Long story. |
| 08:47.50 | seebs | Basically, $dayjob is embedded Linux. |
| 08:48.03 | seebs | So we have people who want to create a complete filesystem for an embedded target. |
| 08:48.17 | seebs | And we have an absolute requirement that you can use the product without having root access to your development machine. |
| 08:48.50 | seebs | So we need to set up something that allows you to manipulate file ownership, create device nodes, and so on... and it doesn't REALLY have to work, but it has to work well enough that tar-type utilities and filesystem image makers can read the intended values. |
| 08:49.19 | ZorbaTHut | aha, interesting |
| 08:49.21 | seebs | We used to use a program called fakeroot, and also one called fakechroot, but we Had Issues, so we wrote a new one called pseudo. |
| 08:49.46 | ZorbaTHut | it doesn't sound *too* tough on Linux, at the very least |
| 08:49.58 | seebs | And it came to pass that I needed to build some portability shims in it, and I decided to do an OS X port on the grounds that this would give me a MUCH better sense of what I was up against for portability. |
| 08:50.10 | seebs | Some of it is pretty straightforward, but some's horrific. |
| 08:50.35 | ZorbaTHut | I assume "put it in a VM" isn't a viable solution |
| 08:50.40 | seebs | You can't *really* intercept syscalls in a practical way (at least, not with livable performance), so you actually have to intercept the syscall wrappers and also any libc functions which make syscalls directly. |
| 08:50.42 | seebs | Right. |
| 08:51.05 | seebs | So you have to do stuff like intercept fopen() so you can peek around in the returned FILE * to figure out what the underlying open() looked like. |
| 08:51.27 | ZorbaTHut | okay, yeah, that sounds like a royal pain :V |
| 08:51.43 | seebs | Example of Hard Problems: |
| 08:51.54 | seebs | Okay, so, you *must* have a server controlling the database, to serialize things. |
| 08:52.21 | seebs | So. It is possible to get a file creation message for a given inode *before* you get the message saying that a file with that inode was removed. |
| 08:52.42 | seebs | The eventual solution I came up with is to add a "deleting" flag, and mark things as "being deleted" before trying to actually delete them. |
| 08:52.58 | seebs | So then if I see something come in with the same inode, I just assume the deletion succeeded, and so on. |
| 08:53.25 | ZorbaTHut | few things are nastier than trying to build a stable API on a badly-defined API :/ |
| 08:53.26 | seebs | What's really amazing is that we've had it in use in the field for three or four years now, and so far we've had one real bug report on it from customers. |
| 08:54.31 | seebs | The reason I love my job: When I proposed to tack three or four weeks onto my already-behind schedule for an experimental replacement of something which absolutely HAD to work for our product to be viable, with intent to give the code away, I got approval from management. |
| 08:55.04 | ZorbaTHut | I must admit that sounds like a really good place to work :) |
| 08:55.32 | seebs | Trion sounds unusually tolerable as game development goes. |
| 08:55.54 | seebs | The closest I've seen to unhappy was a guy who left to work at a startup, and he wasn't so much unhappy as he found something that looked really, really, fun. |
| 08:56.06 | ZorbaTHut | Trion's pretty good too, I'll admit, although I've been blocked on a few of those "hey we really need this gigantic replacement" deals |
| 08:56.37 | seebs | The impression I get is that they are at least willing to *think* about it. Which is pretty cool. |
| 08:57.26 | ZorbaTHut | *nod* yeah, definitely |
| 08:58.39 | seebs | I am liking the proposed progression changes for 1.7. I had been sort of thinking the progression curve looked like it was going exponential, and apparently I was not the only one to notice. |
| 08:59.12 | ZorbaTHut | MMO progression is really tough. I don't think anyone's solved it well |
| 08:59.43 | seebs | I have theories. Actually, I know a guy who presented what looked like convincing math for an argument that you could do decent progression without the underlying assumption of hit point inflation. |
| 08:59.46 | ZorbaTHut | I have some ideas on how to deal with it better, but it's a *significant* redesign of how gearing works |
| 08:59.48 | *** join/#riftuidev Alge (~Alge@124-168-13-181.dyn.iinet.net.au) |
| 08:59.50 | seebs | Yeah. |
| 09:00.00 | *** part/#riftuidev Alge (~Alge@124-168-13-181.dyn.iinet.net.au) |
| 09:00.01 | ZorbaTHut | the problem isn't really "decent progression", it is, as always, making the players happy |
| 09:00.15 | seebs | One thing I really hope is that if Trion ever boosts the level cap, the existing endgame stuff gets retuned for the new cap so it doesn't just disappear and become irrelevant. |
| 09:00.17 | ZorbaTHut | people want to see their numbers increasing noticably |
| 09:00.22 | ZorbaTHut | and that means exponential progression |
| 09:00.32 | seebs | Hmm, good point. |
| 09:00.43 | ZorbaTHut | but simultaneously, people don't want to feel like gear is the only thing that kills bosses |
| 09:00.52 | ZorbaTHut | and that means *no* exponential progression |
| 09:00.57 | seebs | I am still vaguely offended that rift has skill ranks. |
| 09:01.01 | ZorbaTHut | tl;dr: mmo development would be a lot easier if you could somehow get rid of the players |
| 09:01.15 | seebs | I really, really, like CoH's model where skills scale to level without any ranking. |
| 09:01.17 | seebs | Yeah. |
| 09:01.34 | ZorbaTHut | yeah, WoW picked up that model too, and I agree it's really nice. |
| 09:02.01 | seebs | That's my biggest game-engine wishlist item; drop ranks, figure out some other money sink. |
| 09:02.26 | seebs | Well, that, and level scaling/mentoring so that sub-50 zones can be more useful and fun. |
| 09:02.33 | ZorbaTHut | *nod* yeah |
| 09:02.44 | seebs | I would LOVE to be able to take a 50 who skipped a zone and go back to that zone at-level to experience it. |
| 09:03.24 | seebs | Implies a lot of scaling effort, though. |
| 09:03.37 | ShaunO` | i just want the game to be like the end of half life 2 i just wanna pick up the big gun and rape everything with 1 shot |
| 09:03.44 | ZorbaTHut | as a personal thing, I've found I actually have no interest in doing zones at their appropriate level |
| 09:03.59 | ZorbaTHut | I love PvE challenge, and I love questing and plotlines, but I really hate actually *doing* the quests |
| 09:04.15 | seebs | I have done some of them at a level or two above level, or a level or two below. They both produce very different experiences. |
| 09:04.33 | ZorbaTHut | the real problem I have with "a level or two below" is that it's usually the same dance over and over |
| 09:04.44 | seebs | Yeah. There is some of that. |
| 09:04.46 | ZorbaTHut | find a target that won't give you adds --> kill target --> repeat |
| 09:04.58 | seebs | Although since I like stealth builds, I sort of enjoy that. :) |
| 09:05.00 | ZorbaTHut | if the monsters were more interesting I wouldn't mind so much, but the current MMO design has very uninterestnig monsters |
| 09:05.38 | seebs | I have high hopes for Rift pushing a lot of MMO envelopes. I think there's a lot of "you must have this or you can't be in business" crap. |
| 09:05.55 | seebs | But rift is starting to have a lot of stuff which is outside that area, and which provides a lot of interest for me. |
| 09:06.27 | ZorbaTHut | You may also want to look into other Trion MMOs as we make 'em :) |
| 09:06.46 | seebs | Probably not -- two's about my limit, and CoH's social features make it the best for me. :) |
| 09:07.05 | ZorbaTHut | Fair 'nuff :) |
| 09:07.25 | seebs | A friend of mine is really, really, looking forward to Trion's other stuff. Loves the company, just doesn't click with rift. |
| 09:10.14 | seebs | EoN looked pretty PvPish, and the Syfy one, I'm a little scared of the Syfy brand. :P |
| 09:10.36 | ZorbaTHut | The screenshots look gorgeous. I will admit I want to play that game :3 |
| 09:11.03 | ZorbaTHut | I seem to recall that the IP is owned by Trion, not SyFy |
| 09:11.16 | ZorbaTHut | could be wrong on that though |
| 09:11.26 | seebs | One of my friends has been telling me about ToR. The only thing I've liked so far is that you can improve on recipes with practice in the crafting, which sounds sorta cool. |
| 09:11.32 | seebs | Yeah. I am not sure I understand the relationship there. |
| 09:11.48 | ZorbaTHut | from what I hear, ToR is a fantastic singleplayer game with a little bit of co-op functionality and some PvP |
| 09:11.53 | seebs | Yeah. |
| 09:12.27 | seebs | I have been trying to explain to people on the forums that if Trion is hiring 28 people in the office that does Rift, this is probably not a big indicator that the game is about to go under. |
| 09:12.41 | seebs | Sadly, apparently the ability to do math is not valued in our culture. |
| 09:13.13 | ZorbaTHut | from what I've seen watching discussions, every MMO is constantly failing. |
| 09:13.17 | seebs | Yeah. |
| 09:13.49 | seebs | Unless your business people are insane, or you replace every single server every day, or your GMs have mansions, pretty sure Rift has to be making a TON of money. |
| 09:14.25 | ZorbaTHut | I don't know any numbers (and couldn't tell you if I did :P) but I do know that our managers are perpetually in a good mood. Which seems to be a good sign. :) |
| 09:15.18 | seebs | The big killer is that I can't tell participation numbers from subscription numbers. That said, I am seeing more newbies in 1-29 than I used to. |
| 09:15.53 | seebs | I do not think the trial shard thing is a reasonable explanation for people who have no idea what a zone event is. |
| 09:16.27 | seebs | BTW, Trion has a fibber somewhere, apparently; someone was going around saying that the trial shard thing isn't at all like server mergers, it's just preparation for a big promotional thing. |
| 09:16.44 | seebs | A promotional thing big enough to require 42 shards in NA, when the entire active population fits on 16, is Not Realistic. |
| 09:17.29 | ZorbaTHut | That would, I'll admit, be a hell of a promotion. |
| 09:17.49 | seebs | I still think the "trial shard" thing is the least-bad server merger solution I've yet seen. |
| 09:18.06 | seebs | Although they desperately need to increase available slots now. I know people who have 6 characters on each of the three RP shards... |
| 09:44.27 | seebs | ... someone on the forums just explained that a single hardware server would probably have several VMs, each of which would run maybe four shards. |
| 09:44.33 | seebs | So a single machine would be, say, 16 shards. |
| 09:44.46 | seebs | I think this is probably the least accurate description of an MMO's hardware and datacenter I have ever read. |
| 09:45.01 | ZorbaTHut | yeah that's basically backwards |
| 09:45.15 | seebs | I still sorta wonder why rift's server caps seem to be so low. |
| 09:45.50 | seebs | I lean towards a theory that it's game logic, not hardware capacity, in that things like respawn rate get rough on a heavily loaded shard. |
| 09:47.03 | ZorbaTHut | I'll admit that's a factor. Telara isn't all that huge, and there's a definite limit to how many people we can cram into it. |
| 09:47.48 | seebs | A while back, I levelled on a "recommended" shard during a free trial weekend, and with a reasonable spread of people, about 1700-1800 total on the shard... |
| 09:47.54 | seebs | Silverwood was ridiculously hard. |
| 09:48.11 | seebs | If there were two mobs, it was basically impossible to kill them both at once without good AoE. |
| 09:48.13 | ZorbaTHut | and just imagine what it's like on PvP servers :V |
| 09:48.37 | seebs | Yeah. My impression is that if we get a couple more zones here and there, the caps could easily go up. |
| 09:48.57 | seebs | Sadly, the current population spread does not allow me to determine whether the recent hardware upgrades affected population caps. |
| 09:49.17 | seebs | I got lucky back last summer, there were a couple of shards that had a pretty smooth level distribution with not that many 50s on them. |
| 09:50.46 | seebs | My goal is to be obsessive enough about this that within a day or so of releasing any kind of API functionality that would let me run a census, you get told you have to take it out because I now Know Too Much. |
| 09:51.49 | seebs | That said, I would love to have an Inspect.Users({ minLevel = 40, maxLevel = 45}) function. :) |
| 09:52.10 | ZorbaTHut | that wouldn't happen, because I know you exist, and I wouldn't put it in in the first place ;) |
| 09:52.23 | Cairenn | lol |
| 09:52.43 | seebs | It could make sense to make an API that's throttled the way /who is, but lets us be a little more precise about what we're trying to match. |
| 09:52.49 | seebs | The built-in /who is sorta buggy in effect. |
| 09:53.37 | seebs | I suppose in theory Inspect.Users.LoggedIn() wouldn't be game-breaking, but it would conflict with the standard MMO business model of not exposing concrete numbers. |
| 12:11.33 | *** join/#riftuidev Hylax (~philipp@91-66-88-110-dynip.superkabel.de) |
| 12:12.28 | *** part/#riftuidev Hylax (~philipp@91-66-88-110-dynip.superkabel.de) |
| 13:16.13 | *** join/#riftuidev Sunspots (~Sunspots@c213-89-195-33.bredband.comhem.se) |
| 15:44.38 | *** join/#riftuidev Adys (~adys@unaffiliated/adys) |
| 18:49.04 | *** join/#riftuidev ShaunO` (ShaunO@101.98.150.17) |
| 21:31.35 | *** join/#riftuidev seebs (~seebs@74.116.16.132) |
| 21:31.35 | *** mode/#riftuidev [+o seebs] by ChanServ |
| 22:36.23 | *** join/#riftuidev Aurorus (~Aurorus@unaffiliated/aurorus) |
| 23:32.41 | *** join/#riftuidev Torhal_ (~callahan@c-68-62-216-165.hsd1.al.comcast.net) |
| 23:32.41 | *** mode/#riftuidev [+v Torhal_] by ChanServ |